ABSTRACT
OBJECTIVE: Local anatomy and the patient's risk profile independently affect the expansion rate of an abdominal aortic aneurysm. We describe a hybrid method that combines finite element modelling and statistical methods to predict patient-specific aneurysm expansion. METHODS: The 3-D geometry of the aneurysm was imaged with computed tomography. We used finite element methods to calculate wall stress and aneurysm expansion. Expansion rate was adjusted by risk factors obtained from a database of 80 patients. Aneurysm diameters predicted with and without the risk profiles were compared with diameters measured with ultrasound for 11 patients. RESULTS: For this specific group of patients, local anatomy contributed 62% and the risk profile 38% to the aneurysmal expansion rate. Predictions with risk profiles resulted in smaller root mean square errors than predictions without risk profiles (2.9 vs. 4.0 mm, p < 0.01). CONCLUSIONS: This hybrid approach predicted aneurysmal expansion for a period of 30 months with high accuracy.
Subject(s)
Aortic Aneurysm, Abdominal/diagnosis , Aortic Rupture/diagnosis , Aged , Aortic Aneurysm, Abdominal/diagnostic imaging , Aortic Rupture/diagnostic imaging , Aortography/methods , Disease Progression , Female , Finite Element Analysis , Humans , Imaging, Three-Dimensional , Male , Middle Aged , Models, Cardiovascular , Predictive Value of Tests , Prognosis , Radiographic Image Interpretation, Computer-Assisted , Reproducibility of Results , Risk Assessment , Risk Factors , Stress, Mechanical , Time Factors , Tomography, X-Ray Computed , UltrasonographyABSTRACT
Aneurysms of the abdominal aorta enlarge until rupture occurs. We assume that this is the result of remodelling to restore wall stress. We developed a numerical model to predict aneurysm expansion based on this assumption. In addition, we obtained aneurysm geometry of 11 patients from computed tomography angiographic images to obtain patient specific calculations. The assumption of a wall stress related expansion indeed resulted in a series of local expansions, adjusting global geometry in an exponential fashion similar as in patients. Furthermore, it revealed that location of peak wall stress changed over time. The assumptions of this model are discussed in detail in this manuscript, and the implications are related to literature findings.
